Engineering Applications of Integration

Introduction

Integration is the mathematical language of accumulation. In mechatronics engineering, quantities such as displacement, mass, energy, charge, fluid volume, and sensor output are often obtained by accumulating continuously varying rates or distributed properties. A motor’s position is the accumulated velocity; the energy delivered by a variable actuator is the accumulated power; and the mass of a nonuniform component is the accumulated density over its geometry.

This chapter develops definite integration as an engineering tool. The emphasis is not only on evaluating integrals, but also on selecting an appropriate model, identifying assumptions, maintaining units, interpreting the result, and recognizing the limitations of continuous approximations to measured data.

Learning objectives

After completing this chapter, you should be able to:

1. Compute geometric areas and volumes using definite integrals.
2. Determine arc length for planar curves and interpret its assumptions.
3. Calculate work, energy, power, and fluid quantities from variable rates.
4. Find centers of mass and moments for distributed mechanical systems.
5. Compute average values of continuous functions.
6. Approximate integrals from sampled sensor data using numerical methods.
7. Assess units, modeling assumptions, uncertainty, and engineering limitations.

1. Definite integration as accumulation

If a quantity changes at an instantaneous rate r(t), its accumulated change from t=a to t=b is

Here, ΔQ is the total change in quantity Q, r(t) is the rate of change, and t is the independent variable. The integral is applicable when the rate is integrable over the interval. Its units are the units of r multiplied by the units of t.

For example, if velocity is measured in metres per second and time in seconds, then

has units of metres. The signed integral gives displacement. Distance travelled is instead

because distance cannot decrease when the velocity changes direction.

The Fundamental Theorem of Calculus connects accumulation with antiderivatives. If F'(x)=f(x) and f is continuous on [a,b]], then

This theorem is useful when a symbolic antiderivative exists. In engineering practice, numerical integration is often required because data are sampled or the model has no elementary antiderivative.

Units and sign conventions

A reliable integration procedure begins with a sign convention. Positive velocity may mean motion in the chosen positive direction; positive torque may mean counterclockwise rotation; and positive power may represent energy entering a system. A negative result is not automatically an error—it may indicate direction or energy leaving the system.

Dimensional analysis provides a quick check. For instance, force multiplied by distance gives joules:

2. Areas and volumes

Area under a curve

If f(x)≥0 on [a,b]], the area between the curve y=f(x), the x-axis, and the vertical lines x=a and x=b is

If a curve lies below the axis, its definite integral is negative, whereas geometric area is positive. The geometric area between two curves is

or, when f(x) is known to be above g(x) throughout the interval,

The bounds must be determined from the geometry, usually by intersection points or specified dimensions.

Cross-sectional volume

If a solid has cross-sectional area A(x) perpendicular to the x-axis, its volume is

For circular cross-sections of radius r(x),

The disk formula applies when cross-sections are solid circles. For washers with outer radius R(x) and inner radius r(x),

The shell method is useful when cylindrical shells are naturally formed:

For shells parallel to the y-axis, this becomes V=∫_a^b 2πx f(x) dx when the radius is x and the height is f(x). The axis and orientation must be clearly identified before selecting a method.

Worked example 1: volume of a tapered actuator housing

An idealized housing has length 0.40 m and radius varying linearly from 0.030 m at x=0 to 0.050 m at x=0.40. Thus

where x is in metres. The volume is

Expanding and integrating,

The bracketed value is 0.000747 m³, so

The result is approximately 2.35 L. The units are correct because radius squared contributes m² and integration with respect to length contributes m.

3. Arc length and surface generation

For a differentiable planar curve y=f(x) on [a,b]], the arc length is

This follows by approximating the curve with small straight segments. The formula assumes that the curve is sufficiently smooth and that f'(x) is integrable. If x=g(y), an alternative form is

For a parametric curve x=x(t), y=y(t) over t∈[a,b],

Arc length has units of length. It should not be confused with horizontal distance or the integral of the function height.

When a curve y=f(x) is revolved around the x-axis, the surface area is

This assumes f(x)≥0 and no self-intersection. Similar expressions apply around other axes, with radius defined as the perpendicular distance to the axis.

4. Work, energy, and power

Variable force

For a force F(x) acting along a one-dimensional path from x=a to x=b, the work is

This formula assumes that force is collinear with displacement. For a vector force along a path r(t),

where the dot denotes the scalar product. Work is measured in joules.

For rotation, variable torque τ(θ) gives work

where θ is in radians. Radians are dimensionless, so torque in N·m produces work in joules.

Power is the rate of energy transfer:

Therefore, energy delivered over an interval is

For translational systems, P=F·v; for rotational systems, P=τω. These relations require consistent instantaneous force, velocity, torque, and angular velocity.

Worked example 2: variable actuator force

A linear actuator applies F(x)=120+80x N while moving from x=0 to x=0.25 m. The work is

The average force over the displacement is 130 N, and 130 N × 0.25 m also gives 32.5 J. This agreement occurs because the force varies linearly.

5. Fluid quantities and hydrostatic force

Volume flow and accumulated volume

If volumetric flow rate is Q(t) in m³/s, the volume transferred is

If mass flow rate is ṁ(t), accumulated mass is

For constant density ρ, ṁ=ρQ. If density varies, the relationship must be evaluated at each instant: ṁ(t)=ρ(t)Q(t).

Hydrostatic force

A fluid at rest exerts pressure

where p₀ is pressure at the reference surface, ρ is fluid density, g is gravitational acceleration, and h is depth below the surface. The pressure force on a vertical strip of width w(y) and thickness dy is dF=p(y)w(y)dy. Hence

The pressure may be taken as gauge pressure when atmospheric pressure acts on both sides of a surface. This is a static-fluid model; it does not describe dynamic pressure in flowing fluid.

6. Centers of mass and moments

The center of mass identifies the point at which the mass distribution can be represented for translational balance. For a one-dimensional body with linear density λ(x),

and

For a planar lamina with surface density σ(x,y),

For uniform density, σ is constant and cancels in the coordinate ratios. The first moment about the y-axis is M_y=∫∫_R xσ dA, and about the x-axis is M_x=∫∫_R yσ dA.

Worked example 3: nonuniform linear rail

A 2 m rail has density

Its mass is

The first moment about the origin is

Thus

The center lies beyond the midpoint because density increases toward the right end.

For rotational dynamics, the mass moment of inertia about an axis is obtained by integrating squared distance from the axis:

For a linear density distribution along the x-axis about the origin, I=∫_a^b x²λ(x)dx. The units are kg·m². This quantity differs from the first moment and governs angular acceleration through the rotational model τ=Iα when the moment of inertia is constant and external effects are appropriately included.

7. Average values

The average value of a continuous function f(x) over [a,b] is

The formula assumes equal weighting per unit of x. For a time-varying sensor signal, it gives the time average when x=t. If samples are not equally spaced, a weighted or numerical integral must be used rather than a simple arithmetic mean.

For a nonnegative signal, average value is not the same as root-mean-square value. RMS is

RMS is especially relevant to heating and electrical power, whereas the ordinary average describes the mean level.

8. Numerical integration of measured data

Engineering data commonly consist of pairs (t_i,y_i), not a symbolic function. Numerical integration estimates accumulated quantities.

Trapezoidal rule

For equally spaced samples with interval h, the composite trapezoidal rule is

For unequal spacing,

The method treats each interval as a trapezoid. It is generally reliable for smooth data but can be affected by noise, missing samples, offsets, and rapid transients.

Simpson’s rule

For equally spaced data with an even number n of subintervals,

Simpson’s rule can be more accurate for smooth functions, but it requires equally spaced data and an even number of subintervals. It may perform poorly when signals contain sharp discontinuities or substantial high-frequency noise.

Sensor accumulation example

A flow sensor records the following flow rates:

Using the trapezoidal rule with h=2 s,

Thus approximately 6.8 L passed the sensor. The estimate assumes linear variation between samples. If the sensor has a constant bias of 0.05 L/s, the accumulated volume is overestimated by approximately 0.05×6=0.30 L.

9. Applied case: energy estimation for a robotic joint

Background

Consider an illustrative robotic elbow joint monitored during a pick-and-place cycle. A controller records motor power every 0.5 s. Positive power represents electrical energy supplied to the motor; regenerative braking may produce negative power.

Problem

Estimate the net electrical energy supplied during a 2 s cycle using the measurements below.

Analysis and approach

Use the composite trapezoidal rule. Since power is in watts and time is in seconds, the result is joules:

The net supplied energy is therefore approximately 170 J. To estimate energy consumption ignoring recovered energy, integrate only positive power intervals. The intervals give 70, 80, 25, and −5 J respectively; treating the negative interval as recovered energy, positive consumption is approximately 175 J and recovered energy is 5 J.

Outcome and interpretation

The net result and the gross positive-energy result answer different engineering questions. Net energy is appropriate for a battery balance when regenerative energy returns to the battery. Gross positive energy is useful for assessing motor and drive loading. In a real system, conversion efficiencies, sensor calibration, sampling synchronization, and standby power must also be included.

Learning points

1. Integration converts power history into energy.
2. Negative power must be interpreted according to the energy-flow convention.
3. Sampling interval affects numerical accuracy.
4. A physically meaningful result requires system-boundary definition.
5. Calibration and efficiency losses may be as important as numerical integration error.

10. Modeling limitations and best practices

A continuous integral applied to measured data is an approximation. Important limitations include:

• Sampling: Rapid changes occurring between samples may be missed, causing aliasing or underestimation.
• Noise: Random noise can accumulate, while bias produces an error proportional to the integration interval.
• Drift: Integrating an accelerometer signal twice can create large position errors because small offsets accumulate.
• Discontinuities: Piecewise models should be integrated separately over each smooth interval.
• Coordinate and sign errors: Incorrect direction conventions can reverse work, displacement, or torque.
• Model validity: Constant density, rigid-body behavior, negligible friction, or quasi-static pressure may be inappropriate assumptions.
• Numerical precision: Round only after the calculation; retain units and sufficient significant figures.

Good practice is to define the system boundary, record units and sampling times, inspect plots for outliers, compare numerical results with an independent estimate, and report uncertainty or sensitivity when decisions depend on the result. In automated systems, software should flag missing samples, nonmonotonic timestamps, sensor saturation, and implausible accumulated values. AI-based signal reconstruction may help fill gaps, but it must not silently replace measured data; reconstructed intervals should be labeled and validated against physical constraints.

Chapter Summary

Definite integration computes accumulated quantities from continuously varying functions. Areas are obtained by integrating height, volumes by integrating cross-sectional area, and arc lengths by integrating differential path length. Variable force and torque produce work through spatial or angular integration, while power integrates over time to give energy. Flow rate and mass flow rate integrate to volume and mass. Distributed density leads to mass, centers of mass, first moments, and moments of inertia. Average values are normalized integrals, and measured sensor data can be accumulated using trapezoidal or Simpson’s rules. Correct engineering use requires appropriate bounds, sign conventions, assumptions, units, numerical methods, and awareness of sampling and sensor errors.

Key Terms

Definite integral: An integral evaluated between specified limits.

Accumulation: Total quantity obtained by integrating a rate or density.

Displacement: Signed change in position obtained by integrating velocity.

Distance: Total path length, obtained by integrating speed.

Cross-sectional area: Area of a slice used to calculate volume.

Arc length: Length of a curve.

Work: Energy transferred by a force or torque through displacement or rotation.

Power: Rate of energy transfer.

Volumetric flow rate: Volume transferred per unit time.

Mass flow rate: Mass transferred per unit time.

Hydrostatic pressure: Pressure in a fluid at rest due to depth and reference pressure.

Center of mass: Location representing the mass-weighted mean position.

First moment: Integral of position multiplied by mass or area density.

Moment of inertia: Integral of squared distance multiplied by differential mass.

Average value: Integral divided by the length of the independent-variable interval.

Trapezoidal rule: Numerical integration method based on trapezoidal segments.

Simpson’s rule: Numerical integration method using weighted parabolic approximations.

Sampling interval: Time or spatial separation between measured data points.

Bias: Persistent measurement offset that accumulates under integration.

System boundary: Deliberately defined region over which a balance is performed.

Review Questions

Short Answer Questions

1. Distinguish displacement from distance when velocity changes sign.
2. State the disk and washer formulas for volume and identify their assumptions.
3. Why must angular displacement be expressed in radians in rotational work calculations?
4. Define the center of mass using a linear density function.
5. What is the principal difference between the trapezoidal rule and Simpson’s rule?

Long Answer Questions

1. Explain how definite integration is used to calculate work, energy, and power, including units and sign conventions.
2. Derive the expression for the arc length of a differentiable curve from small line segments.
3. Explain how hydrostatic force on a vertical plate is obtained by integrating pressure over differential strips.
4. Compare symbolic integration with numerical integration for engineering sensor data.
5. Discuss how density variation affects mass and center-of-mass calculations.

Analytical / Application Questions

1. A motor’s velocity is v(t)=0.8t²−2t m/s for 0≤t≤3. Calculate its displacement and total distance over the interval.
2. A pump flow rate is Q(t)=4+0.5t L/s for 10 s. Calculate the transferred volume and explain the unit conversion if the answer is required in cubic metres.
3. A plate has linear density λ(x)=3+x kg/m over 0≤x≤2. Determine its mass and center of mass, and explain why the center is not at the midpoint.

Multiple Choice Questions with Answers

Question 1

A velocity function is integrated with respect to time to obtain which quantity?

A. Acceleration

B. Displacement

C. Force

D. Power

Correct Answer: B. Displacement

Question 2

The volume of a solid with cross-sectional area A(x) perpendicular to the x-axis is given by which expression?

A. ∫ A(x) dx

B. ∫ A′(x) dx

C. ∫ xA(x) dx only

D. ∫ √A(x) dx

Correct Answer: A. ∫ A(x) dx

Question 3

A force F(x) acts along a one-dimensional path. Which formula gives its work from x=a to x=b?

A. ∫ F′(x) dx

B. ∫ F(x)/x dx

C. ∫ F(x) dx

D. F(b)−F(a)

Correct Answer: C. ∫ F(x) dx

Question 4

Power is measured in watts. Integrating power with respect to time gives a quantity measured in:

A. Newtons

B. Joules

C. Pascals

D. Kilograms per second

Correct Answer: B. Joules

Question 5

A flow rate of 2 L/s is maintained for 5 s. What volume is transferred?

A. 0.4 L

B. 2.5 L

C. 7 L

D. 10 L

Correct Answer: D. 10 L

Question 6

For a linear density λ(x), which expression gives the center-of-mass coordinate x̄?

A. ∫λ(x)dx

B. (∫xλ(x)dx)/(∫λ(x)dx)

C. ∫x/λ(x)dx

D. (∫λ(x)dx)/(∫xλ(x)dx)

Correct Answer: B. (∫xλ(x)dx)/(∫λ(x)dx)

Question 7

The average value of f(x) on [a,b] is:

A. ∫_a^b f(x)dx

B. (b−a)∫_a^b f(x)dx

C. (1/(b−a))∫_a^b f(x)dx

D. f(b)−f(a)

Correct Answer: C. (1/(b−a))∫_a^b f(x)dx

Question 8

Using the trapezoidal rule with h=1, data values f₀=2, f₁=4, and f₂=6 give an approximate integral of:

A. 4

B. 6

C. 8

D. 12

Correct Answer: C. 8

Question 9

For a curve y=f(x), which expression gives its arc length on [a,b]?

A. ∫_a^b f(x)dx

B. ∫_a^b √(1+[f′(x)]²)dx

C. ∫_a^b f′(x)dx

D. ∫_a^b [f(x)]²dx

Correct Answer: B. ∫_a^b √(1+[f′(x)]²)dx

Question 10

A constant sensor bias in a rate measurement generally causes which effect after integration?

A. An accumulated error proportional to the integration interval

B. No error because integration removes bias

C. An error independent of time

D. Only a change in measurement units

Correct Answer: A. An accumulated error proportional to the integration interval